Skip to main content

Self-hosted agent memory

Your AI agents finally share the same brain.

Give your coding tools the same local memory layer. Persistent context, reusable recall, and shared knowledge without another hosted dependency.

Windows download available now. macOS and Linux coming soon.

agent-brain://memorylocal
brain_remember({
  type: "decision",
  project: "checkout",
  content: "Use Stripe webhooks for fulfillment",
})

brain_recall("how do we ship purchases?")
// retrieves the decision in Claude, Codex, Cursor, or REST
40-70%
token reduction from semantic cache
57ms
P95 recall latency in benchmarks
8
MCP tools for shared memory
$0
community tier, self-hosted

What it does

Durable project context without another hosted memory silo.

Agent Brain gives your local agents a shared memory store, semantic cache, and governance layer that outlives any single chat window.

M

Persistent memory

Agents remember durable decisions, patterns, fixes, preferences, and project context across sessions.

C

Semantic cache

Similar prompts can reuse prior answers instead of burning tokens again.

N

Cross-agent context

Supported coding tools, scripts, and local integrations can all connect to the same shared memory.

L

Local model brain ops

Core memory operations run on hardware you control.

A

Memory governance

Audit, deduplicate, archive, restore, and consolidate memories so recall stays relevant as the corpus grows.

D

Desktop control plane

The desktop app manages setup, health, knowledge exploration, and monitoring.

How it works

Local services, one shared context layer.

1

Start the local stack

The desktop app or source setup starts the local services Agent Brain needs.

2

Connect your agents

Connect your agents through the supported local endpoints.

3

Store durable knowledge

Agents store durable project knowledge with structure and scope.

4

Recall what matters

Future sessions retrieve relevant context and share discoveries across tools.

Local memory store
Fast recall layer
Shared agent context
Governance workflows
Local model support
MCP + REST adapters

Adapter layer

MCP clients use localhost:3100/mcp. REST and WebSocket clients use localhost:9090/api/v1.

Comparison

Built for coding-agent workflows.

Self-hosted

Agent BrainYesmem0NoGPTCacheNoLettaNo

Semantic cache

Agent BrainYesmem0NoGPTCacheYesLettaNo

Cross-agent sharing

Agent BrainYesmem0NoGPTCacheNoLettaNo

Local model operations

Agent BrainYesmem0NoGPTCacheNoLettaNo

Desktop stack control

Agent BrainYesmem0NoGPTCacheNoLettaYes

MCP support

Agent BrainYesmem0YesGPTCacheNoLettaNo

Agent-agnostic adapters

Agent BrainYesmem0NoGPTCacheNoLettaNo

Community tier

Agent BrainYesmem0NoGPTCacheYesLettaYes

Pricing

Start free. Upgrade when you need more.

Community is free to use. Pro adds unlimited memories, semantic cache, and desktop backup for commercial workflows.

Community

$0

7-day free trial. All features included. No credit card required.

  • YesUp to 250 stored memories
  • YesUp to 3 connected agents
  • YesMCP and REST adapters
  • YesDesktop app
  • YesLocal model support
  • YesCommunity support (Discord)
Download Free

Pro

$19/mo

For individual developers doing commercial work. Single-user license.

  • YesUnlimited stored memories
  • YesUnlimited connected agents
  • YesSemantic cache enabled
  • YesDesktop backup and restore
  • YesCommercial use permitted
  • YesPriority email support (next business day)

License delivered by email

Enterprise

From $4,800/yr

For organizations. Custom quote — contact us.

  • YesEverything in Pro
  • YesUp to 20 seats
  • YesDedicated onboarding call
  • Yes4-hour SLA (business hours)
  • YesFirst access to SSO, RBAC, compliance pack
Contact Sales

FAQ

Common questions

Does Agent Brain send my code or memories to a cloud service?

No. Agent Brain is self-hosted. Core memory and recall operations run on your machine or infrastructure.

Which agents can use it?

Agent Brain works with supported coding tools, scripts, and local integrations through its public interfaces.

What does the semantic cache do?

It checks whether a similar query already has a useful answer. If the similarity score clears your threshold, agents can reuse the cached response instead of spending fresh tokens.

Do I need Docker?

Yes for the packaged local stack. Docker is required for the local services Agent Brain manages.

Is Pro checkout live?

Yes. Click 'Get Pro' on the pricing page to subscribe monthly ($19/mo) or annually ($190/yr). Community downloads are available now from the download page.

Can I use Community for commercial work?

No — Community is for personal and non-commercial use only. If you are using Agent Brain as part of paid work, a business, or a team, you need a Pro or Enterprise license.

Affiliate Program

Promote Agent Software Suite. Earn 30% recurring.

Refer paying builders and earn 30% on every invoice they pay for the next 12 months.

30%

Commission

12 invoices

Window

60 days

Cookie

Set up Agent Brain and connect your tools.

Start the local stack, point MCP at localhost, and let every agent remember the same durable context.